The section Farinosae was first denned by Pax (1) in 1889 with citation of P. farinosa and P. sibirica as representative species. He published at the same time his section Auriculatae. The only distinction given of apparent moment between the two sections is the shape and size of the capsule. In Auriculatae it is globose and included in the calyx, while in Farinosae it is cylindric and more or less exserted. This diagnostic character was found to be unsatisfactory, with the result that in 1905 Pax (2) in his Monograph combined the two sections under Farinosae. There was still an undercurrent of opinion that a division was justified, and at the Third Primula Conference in 1913 Bayley Balfour (3) reverted to the previous arrangement without, however, any stated reasons. At the Fourth Primula Conference held in 1928 the sections of the genus were reviewed by Smith and Forrest (4), who again subordinated Auriculatae to Farinosae. In 1932 appeared Bruun's Cytological Studies in Primula (5), with a detailed analysis of the species then available in cultivation. Here valid evidence was given that certain species assessed in the past as within Auriculatae did differ cytologically from the general run of species included without question in the Farinosae. But the cytological evidence taken alone tends to associate these Auriculatae with other species quite dissimilar from the broad morphological point of view. The problem is referred to again by Wright Smith (6) in his Hooker Lecture, where it is admitted that the cytological data do support the view that there is a degree of fundamental difference between the two sections. But the difficulty still remains of finding satisfactory macroscopic distinctions. The present authors have come to the conclusion that it is better to keep both under the general heading Farinosae rather than to separate them by slender and often misleading characters supported as these may be by the cytology. In any case much remains to be done on the cytological side, as the majority of the species are yet to be analysed, and for general purposes the incorporation of the cytological data in an analytical key would not be satisfactory.